We would want the solution to the safety problem before somebody figures out the solution to the AI problem.

The greatest existential risks over the coming decades or century arise from certain, anticipated technological breakthroughs that we might make in particular, machine super intelligence, nanotechnology and synthetic biology. Each of these has an enormous potential for improving the human condition by helping cure disease, poverty, etc. But one could imagine them being misused, used to create powerful weapon systems, or even some kind of accidental destructive scenario, where we suddenly are in possession of some technology that's far more powerful than we are able to control or use wisely.

Our approach to existential risks cannot be one of trial-and-error. There is no opportunity to learn from errors. The reactive approach - see what happens, limit damages, and learn from experience - is unworkable. Rather, we must take a proactive approach. This requires foresight to anticipate new types of threats and a willingness to take decisive preventive action and to bear the costs (moral and economic) of such actions.

We would want the solution to the safety problem before somebody figures out the solution to the AI problem.

Human nature is a work in progress.

Knowledge about limitations of your data collection process affects what inferences you can draw from the data.

The Internet is a big boon to academic research. Gone are the days spent in dusty library stacks digging for journal articles. Many articles are available free to the public in open-access journal or as preprints on the authors' website.